Chan Mo-Che
Chan Mo-Che
Cornernet_Lite is an enhancement of Cornernet with the same backbone network. Cornernet_Lite has better inference speed and higher mAP than Cornernet. Centernet uses most parts of backbone network of Cornernet....
After reviewing CenterNet codes and CornerNet_Lite codes, to implement the centernet_inference(db, nnet, image, decode_func=decode) function is likely necessary to use CenterNet in my project. https://github.com/princeton-vl/CornerNet-Lite/blob/master/core/test/cornernet.py#L75 Due to match necessary interfaces,...
My result also shows that Yolov3 has better accuracy than Cornernet_Lite. I have no idea why the paper shows that Cornernet_Lite has higher mAP.
@heilaw Sure. I never evaluated Cornernet_Lite on COCO. I just train my own dataset to detect license plates in order to evaluate the possibility to replace my current object detection...
Same with ubuntu 18.04. And mktndn's solution does not work in ubuntu 18.04. Any other suggestion for ubuntu 18.04? Thank you.